There is one fine-grained functional sequence in morpho-syntax: Japanese complex verbs are like Slavic prefixes Yoshio Endo Kanda University

Bartosz Wiland University of Poznań

BCGL 7: The Morphology-Syntax Interface 17-18 Dec 2012

1. Synopsis We report on a robust symmetry between the Polish and the Japanese sequence of projections in what superficially looks like belonging to different domains namely • morphology: verbal prefixes in Polish (and Slavic, more generally) • syntax: multiple verbs in Japanese General claim: This parallellism strongly supports the thesis that there exists one and only functional sequence of heads (fseq) in morphosyntax which is invariantly ordered by UG (eg. Cinque (1999)) and the surface differences among particular languages in the order of elements that instantiate this fseq result solely from movement (not from the variation in fseq itself). Specific thesis: The sequence of syntactic heads behind the multiple verbal prefixes in Polish (and in other Slavic languages) and the multiple verb system in Japanese instantiates the same subset of fseq, with primitives such as 'distributive', 'saturative', 'excessive', etc. heading their own projections in the syntax of both Polish and Japanese. For instance:1 (1)

DIST po > SAT na (Polish) a. po-na-jadać się owoców DIST- SAT-eat self fruits 'to eat some fresh fruits (to the point of satisfaction)' b. * na-po-jadać się SAT- DIST-eat self

(2)

DIST nare < SAT makur (Japanese) a. Merii-ga posutaa-o hari-makuri-nare-tei-ru. Mary-NOM poster- ACC paste-roll.up-get.used.to-ASP-PRESENT ‘I is used to pasting up posters to the point that there is no poster left’ b. * Merii-ga posutaa-o hari-nare-maku(r)- tei-ru. Mary-NOM poster- ACC paste-get.used.to-roll.up-ASP-PRESENT

Roadmap: • fseq behind multiple prefixation facts in Polish • the ordering of Japanese complex verbs and argue that the reversed order of verbs wrt. 1 We adopt the following abbreviations: ATT - attenuative, COMPL - completive, CUML - cumulative, DELIM - deliminative, DIST - distributive, EXC - excessive, PERD - perdurative, REP - repetitive, SAT saturative, TERM - terminative

-1-

fseq behind the Polish prefixes results from the roll-up derivation of the Japanese clause • partial roll-up of the fseq in Bulgarian prefixation

2. Functional sequence of Polish verbal prefixes Slavic prefixes form two heterogeneous classes wrt. their syntactic and semantic properties (Ramchand (2004), Romanova (2004), Svenonius (2004b), DiSciullo and Slabakova (2005), among others): • superlexical (or vP-external) prefixes -- contribute aspectual meanings • lexical (or vP-internal) prefixes -- argument structure of the predicate Verbal prefixes in Polish: • multiple prefixes • format for stacked prefixes: [ super-lex [ super-lex [ lex [ V ]]]] • Wiland (2012): vP-external prefixes2 in Polish lexicalize the articulate fseq of aspectual heads on top of the vP:3 (3)

[ Dist0 [ Att0 [ Delim0 [ Cuml0 [ Sat0 [ Perd0 [ Exc0 [ Rep0 [ Compl0/Term0 [ v0 [ ... • aspectual notions such as 'distributivity', 'saturation', or 'repetition' are heads of syntactic projections spelled-out as prefixes on top of the verb stem • only certain orders of stacked prefixes are well-formed, eg.:

(4)

a.

po-prze-bijać oferty DIST- EXC-hit offers 'to make better offers (several times)' b. * prze-po-bijać EXC-DIST-hit

(5)

a.

(6)

a.

(7)

a.

na-do-krajać chleba SAT- COMPL-cut bread 'to slice even more bread' b. * do-na-krajać COMPL-SAT-cut po-prze-pisywać DELIM-REP-write 're-write something (a little bit)' b. * prze-po-pisywać REP-DELIM-write na-prze-siadywać się DIST- PERD-sit self 'to sit for a long time (in some places on numerous occasions, e.g. in pubs)'

2 In this work we discuss only the superlexical/vP-external prefixes and do not adopt or make any claims about

the class of lexical/vP-internal prefixes. 3 The immediately pre-verbal head position is indicated in (3) as Compl0/Term0 since, as noted in Wiland (2012), the data analyzed there is insufficient to determine their ordering with respect to each other. The exact ordering of Compl(etive)0 and Term(inative)0 with respect to each other is not essential to what follows.

-2-

b. * prze-na-siadywać się PERD-DIST-sit self Wiland (2012): generalization about Polish aspectual prefixes: • given any two vP-external prefixes that can stack in the order X>Y, the reversed order Y>X is ill-formed (holds without without exception) • only multiple prefixation patterns that observe the fseq in (3) are well-formed, eg.: (8)

DIST po > ATT pod a. po-pod-duszać (mięso w garnku) DIST- ATT-stew (meat in pot) 'to stew meat in a pot a little bit' b. * pod-po-duszać ATT- DIST-stew

(9)

DIST po > SAT na a. po-na-jadać się owoców DIST- SAT-eat self fruits 'to eat some fresh fruits (to the point of satisfaction)' b. * na-po-jadać się SAT- DIST-eat self

(10) DIST po > EXC prze a. po-prze-krzykiwać innych DIST- EXC-shout others 'to shout louder than others' b. * prze-po-krzykiwać EXC-DIST-shout (11) DIST po > REP prze a. po-prze-pisywać nuty na pięciolinię DIST- REP-write tunes onto stave 'to re-write tunes onto a stave' b. * prze-po-pisywać nuty REP-DIST-write tunes (12) DIST po > PERD prze a. po-prze-siadywać w knajpach całe dni DIST- PERD-sit in pubs all days 'to spend all days in pubs' b. * prze-po-siadywać PERD-DIST-sit (13) DIST po > COMPL do a. po-do-kańczaj swoje DIST- COMPL-finish your 'finish your assignments' b. * do-po-kańczaj COMPL- DIST-finish

zadania assignments

(14) CUML na > PERD prze a. na-prze-pisywać się listów CUML-PERD-write self letters -3-

'to re-write letters (in bulk, for a long time)' b. * prze-na-pisywać się PERD-CUML-write self (15) CUML na > TERM od a. na-od-rabiać się zadań domowych CUML-TERM-do self homework assignments 'to do a lot of homework assignments' b. * od-na-rabiać się TERM-CUML-do self fseq in (3) also explains a constraint on stacking syncretic super-lexical prefixes in Polish: • syncretic prefixes na- and prze- do not stack together • na- (syncretic for cumulative and saturative) can stack with other prefixes but not with one another: (16) a. po-na-krajać DIST- CUML-cut ‘to cut extensively’ b. po-na-pijać się DIST- SAT-drink self ‘to drink to the full’ (17) a. * na-na-krajać b. * na-na-pijać się • prze- (syncretic for repetitive, excessive and perdurative) can stack with other prefixes but not with one another: (18) a. po-prze-rabiać DIST- REP-make ‘to remake’ b. po-prze-krzykiwać DIST- EXC-shout ‘to outshout’ c. na-prze-siadywać DIST- PERD-sit ‘to sit for a long time’ (19) a.* prze-prze-rabiać b.* prze-prze-krzykiwać c. * prze-prze-siadywać • this contrasts with syncretic distributive and deliminative po-, which can be stacked on one another (often when the verb stem is preceded by a lexical prefix): (20) a. po-po-w-kładać DIST- DELIM-in-put ‘to put something in’ b. po-po-w-nosić DIST- DELIM-in-bring ‘to bring something in’ -4-

c. po-po-w-klejać DIST- DELIM-in-paste ‘to paste something in’ Claim made in Wiland (2012): underspecification approach to prefix syncretism in Polish is untennable -- syncretic prefixes either denote considerably different semantic concepts (visible especially on the example of prze-), different syntactic behavior, or both, eg: • saturative na- (measure functor, abundance reading) demands a reflexive clitic, while cumulative na- (collectivizing reading) does not: (21) Saturative naa. Dzieci na-jadły *(się) makaronu. children SAT-eat self pasta 'Children have eaten up pasta (to the point of complete fullness).' b. Kibice po-na-pijali *(się) piwa. fans DIST- SAT-drank self beer. 'The soccer fans have drank up beer (to the point of complete satisfaction).' (22) Cumulative na-4 a. Ich dzieci na-brały (*się) za dużo rzeczy na wakacje. their children CUML-collect self too much things on holidays 'Their children have taken too many things for holidays.' b. Dzieci na-zabierały (się) poziomek w lesie. children CUML-pick.up (self) wild strawberries in forest 'Children have picked up (a lot of) wild strawberries in the forest.' Alternative there: syncretism as overspecification (as in Caha (2009) on case syncretism): • syncretism of na- and prze- is restricted to adjacent positions in the resulting fseq in (3) and to non-adjacent positions in the case of syncretic po- : (23) [Dist0 [Att0 [Delim0 [Cuml0 [Sat0 [Perd0 [Exc0 [Rep0 [Compl0 [Term0 [... | | | | | po pod po na prze do od

3. Japanese complex verbs The surface sequence of Japanese multiple verbs -- called V-V compounds in the literature (cf. Kageyama (1989, 1993, 2009) and Nishiyama (2008)) -- constitutes a mirror-image of the fseq in (3) found in Polish: (16) a. John-ga hon-o yomi-hazime-ta. John-NOM book-ACC read-begin- PAST ‘John began to read a book.’ b. John-ga Bill-o osi-taosi-ta. John-NOM Bill-ACC push-topple- PAST ‘John pushed Bill down.’

4 In cases like in (22b) where cumulative na- can co-occur with a reflexive clitic, it becomes subject-oriented,

very much like saturative na- is. Cumulative na- without a reflexive clitic is object-oriented.

-5-

• previous literature: V-V compounds (two verbs) as in (17), constrained in various ways (cf. Kageyama (1993), Matsumoto (1998) and Yumoto (2005)): (17) DIST nare5 a. John-ga uta-o utai-nare-tei-ru. John-NOM song-ACC sing-get.used.to-ASP-PRESENT ‘John is used to to singing a song.’ EXC makas b. John-ga Merii-o utai-makasi-tei-ru. John-NOM Mary-ACC sing-exceed -ASP-PRESENT ‘John is outsinging Mary.’ • complex verbs -- the second verb of the V-V compounds may be further followed by another V to the effect that the compound comprises three verbs, as in (18a), and the reversed order of the second verb and the third verb is ill-formed, as in (18b):6 (18) DIST nare < EXC makas a. John-ga Merii-o utai-makasi-nare-tei-ru. John-NOM Mary-ACC sing-exceed-get.used.to-ASP-PRESENT ‘John is used to to outsinging Mary’ b. * John-ga utai-nare-makasi-tei-ru. John-NOM sing-tend-exceed- ASP-PRESENT • complex verb that observe the mirror-image of the fseq of Polish prefixes in (3) are well-formed, all reverse orders are ill-formed, as in the (b) examples: (21) DIST nare < SAT makur a. posutaa-o hari-makuri-nare-tei-ru. poster- ACC paste-roll.up-get.used.to-ASP-PRESENT ‘I am used to pasting up posters to the point that there is no poster left.' b. * posutaa-o hari-nare-maku(r) -tei-ru. poster- ACC paste-get.used.to-roll.up-ASP-PRESENT

5 One may wonder if the second verb nare ‘get.used.to’ really heads the verbal complex utai-nare ‘sing-

get.used.to‘. The following facts seem to indicate the headedness of nare ‘get.used.to’. The fist verb utai ‘sing’ is atelic and is, thus, compatible with for-adverbials like 3-zikan ‘for three hours’ as in (ia). In contrast, the second verb nare ‘get.used.to’ is telic and is, thus, compatible with in-adverbials like 3-zikan-de ‘in three hours’ as in (ib). With this distinction in mind, consider (ic), where the verbal complex utai-nare ‘sing-get.used.to‘ is compatible with an in-adverbial, not with a for-adverbial. This indicates that the right-most verb heads the complex verbal constituent. (i)

a.

John-ga 3-zikan uta-o utat-ta. John-NOM 3-hour song-ACC sing-PAST ‘John sang a song for three hours.’ b. John-ga uta-ni nare-ta. John-NOM song-ACC get.used.to-PAST ‘Lit. John get used to a song in three hour s.’ c. John-ga *3-zikan/3-zikan-de uta-o utat-nare-ta. John-NOM 3-hour/3hour-in song-ACC sing-get.used.to PAST ‘John sang a song *for three hours/in three hours.’

6 Unlike in the domain of Slavic prefixes, we are providing approximate lexical meanings of constituents

inside the Japanese complex verbs denoting more abstract notions such as 'excessive', 'complete', etc., whenever it is possible. For this reason makas can be most naturally translated as 'exceed', ager as 'complete', etc.

-6-

(22) DIST nare < CUML tamer a. miuz-o tori-tame-nare-tei-ru. water-ACC collect-save-get.used.to-ASP-PRESENT 'I am used to collecting water.' b. * mizu-o tori-nare-tame-tei-ru. water-ACC collect-get.used.to-save-ASP-PRESENT (23) DIST nare < EXC makas a. John-o tabe-makasi-nare-tei-ru. John-ACC eat-defeat-get.used.to-ASP-PRESENT 'I am used to outeating John.' b. * John-o tabe-nare-makasi-tei-ru. John-ACC eat-get.used.to-defeat-ASP-PRESENT (24) DIST nare < REP naos7 a. arubaito-de posutaa-o hari-naosi-nare-tei-ru. part.time-as poster-ACC paste-do.again-get.used.to-ASP-PRESENT ‘I am used to to re-pasting posters as a part-time.' b. * arubaito-de posutaa-o hari-nare-naosi-tei-ru. part.time-as poster- ACC paste- tend to re-pasting-ASP-PRESENT (25) DIST nare < COMPL ager a. sigoto-de takusanno kimono-o nui-age-nare-tei-ru. Job-as many kimono- ACC sew-complete-get.used.to-ASP-PRESENT 'I am used to to sewing many kimonoes for my job.' b. * sigoto-de takusanno kimono-o nui-nare-age-tei-ru. Job-as many kimono-ACC sew-get.used.to-complete-ASP-PRESENT • like in Polish prefixes, this pattern holds not only with the distributive verb nare- but with other complex verbs, eg.: 7 As we saw in footnote 6, the right-most verb heads the verbal complex and determines the compatibility in

aspect between for-adverbials and in-adverbials. It is predicted that in the case where a verb is followed by two auxiliary verbs, the second verb determines the compatibility between for-adverbials and in-adverbials. The prediction is borne out in the case in (24) in an interesting way. Here, the fist verb hari ‘paste’ and the second verb naosi ‘do.again’ are ambiguous between for-adverbials and in-adverbials, as in (ia-b). With the foradverbial 3-zikan ‘for three hour’, the sentence in (ia) means that the duration time of John’s poster-pasting work continued for three hours. In contrast, with the in-adverbial 3-zikan-de ‘in three hour’, the sentence in (ib) means that John finished pasting posters in three hours. (i)

a.

John-ga 3-zikan/zikan-de posutaa-o hat-ta John- NOM 3-hour/3-hour-in poster-ACC paste-PAST ‘John pasted posters for three hours/in three hours.' b. John-ga 3-zikan/zikan-de posutaa-o naosi-ta John- NOM 3-hour/3-hour-in poster-ACC do.again-PAST ‘John changed posters for three hours/in three hours.'

Recall from footnote 6 that the third verb nare ‘get.used.to‘ is not compatible with for-adverbials but with inadverbials. If the right-most verb heads the whole verbal complex and determines the aspectual concord between for-adverbials and in-adverbials, then it is predicted that the whole verbal complex in (24) hari-naosinare ‘paste-do.again-get.used.to‘ should only be compatible with the in-adverbial 3 zikan-de ‘in three hours‘, not with the for-adverbial 3-zikan ‘for three hours‘. The prediction is borne out as in (ii) below: (ii)

a.

3-zikan-de/*3-zikan posutaa-o hari-naosi-nare-ta. 3-hour-in/*3-hour poster-ACC paste-do.again-get.used.to-PRESENT ‘I got used to to re-pasting posters in three hours.'

-7-

(26) CUML tamer < REP naos a. kitte-o arubamu-ni hari-naosi-tame-tea-ru. stamp-ACC album-DAT paste-do.again-save-ASP-PREESENT 'I have re-pasted up stamps to the albums (in bulk).' b. * kitte-o hari-tame-naosi-tea-ru. stamp-ACC paste-save-do.again-ASP-PRESENT (27) CUML tamer < COMPL ager8 a. sokoni takusano kimono-o nui-age-tame-te-aru. there many kimono-ACC sew-complete-save-ASP-PRESENT 'There are many kimonoes sewed up kimono (in bulk) out there.’ b. * sokoni takusano kimono-o nui-tame-age-te-aru. there many kimono-ACC sew-complete-save-ASP-PRESENT • another paralellism with Polish prefixes: inceptive hazime(r) appears in the rightperiphery in the V-V-V sequence: (28)

Sekken-o tukai-oe-hazimeru. soap-ACC use-finish-begin ‘I will start to use up soap.’

• Note that Polish lexical/VP-internal inceptive prefix za- is not part of fseq in (3); inceptive verb hajimeru is the right-most element inside the verb -- this part of structure does not undergo roll-up (ie. as a bottom layer over which all lower nodes move up)

4. Surface morpheme order by roll-up in Japanese So far: Japanese multiple verbs instantiate the same fine-grained fseq as Polish prefixes, with primitives such as 'excessive' or 'cumulative', etc. heading their own projections, but in a reversed order. This order is derived by (upward) roll-up, typical to Japanese clause. Clausal functional particles in Japanese reflect the morphological make up of Polish verbs: (29) narabe-rare-tei-ta-yooda arrange-Passive-Aspect-Tense-Mood ‘(Things) seem to have been arranged.’ • according to the Mirror Principle, this sequencing of the particles spells out the hierarchy of heads is follows: (30) [verb stem]
‘do.again’. More study is required into the nature of this verb. 9 Syntactically speaking, the Theme Vowel that merges directly with the root in Polish (and in all Slavic) has been identified as v0 or some variants of it in Jabłońska (2007) and Wiland (2009). What is descriptively important here is just that it merges with the root. Also, as it is well known, the phonetically null exponent of the Active Voice morpheme has its overt Passive counterpart -n-/-t-/-on- as, for example, in rob-i-on-y 'done'.

-8-

• surface word order in Japanese is derived by the roll-up derivation (cf. Cinque (2006), Aboh (2003) and especially Koopman (2005)): (32) a. [A [B [C…]]], where A asymmetrically c-commands BC, and B asymmetrically c-commands C. b. [A [[C…][B

]]] ==> [[[C…][B ]][A

]]

• in such a roll-up derivation, verbs suffixed by several clausal functional heads in Japanese are derived in the following way (cf. Endo (2007) for roll-up movement in Japanese): (33) Surface order: Base order: Derivation:

V-Voice-Aspect Aspect-Voice-V Aspect-Voice-[V] Aspect-[V-Voice] [V-Voice-Aspect]

move V over Voice → move [V-Voice] over Aspect →

Extending this logic to fseq in (3), repeated below as (34): Polish spells it out without movement as pre-verbal prefixes and Japanese spells it out as heads of full verbs in a roll-up derivation: (34) [ Dist0 [ Att0 [ Delim0 [ Cuml0 [ Sat0 [ Perd0 [ Exc0 [ Rep0 [ Compl0/Term0 [ v0… • illustration of rolling-up of a sequence (consistent upon) (3)/(34): (35) DIST nare > SAT makur a. Merii-ga posutaa-o hari-makuri-nare-ta. Mary-NOM poster- ACC paste-roll.up-get.used.to- PAST ‘I got used to pasting up posters to the point that there is no poster left’ b. * hari-nare-maku(r)-ta. paste-get.used.to-roll.up- PAST c. …[ DIST nare ‘get.used.to'…[SAT makuri ‘roll.up’ [ v0 hari ‘paste’ move v0 over SAT → …[ DIST nare ‘get.used.to‘…[v0 hari ‘paste’ [SAT makuri ‘roll.up’… move [v0 [SAT… ]] over DIST → …[ v0 hari ‘paste’ [SAT makuri ‘roll.up’]] [DIST nare ‘get.used.to‘]…

5. Partial roll-up in Bulgarian Unlike in Polish, which can stack up to two superlexical/vP-external prefixes at the same time, in Bulagarian a verb may host up to 7 prefixes (though, as reported in Istratkova (2004), combinations of more >4 prefixes are infrequent). The surface order of Bulgarian prefixes reported in Istratkova (2004) is as in (39): (36) a. iz-po-na-pro-dam COMPL- DIST- CUML-through-sell ‘sell completely a lot of things one by one’ -9-

b. *na-iz-po-pro-dam, *na-po-iz-pro-dam, etc. (37) a. za-iz-po-nareżdam INCP-COMPL-DIST-arrage ‘start arranging’ b. *iz-po-za-nareżdam, *iz-za-po-nareżdam (38) a. iz-po-raz-każa COMPL- DIST- EXC-narrate ‘narrate completely one by one’ b. *raz-iz-po-każa (39) The surface order of superlexical prefixes in Bulgarian: ATT

po

> INCP > TERM > COMPL > DIST > CUML > EXC > REP za do iz po na raz pre

The order in (39) considerably differs from the fseq of Polish prefixes in (3); however: • Istratkova (2004): the prefixes inner to the verbal stem are derived lower as they fall in the scope of the outer prefixes • proposal: Bulgarian prefixes are base-generated in an order which is consistent upon the same fseq as in Polish, ie: (40)

• applying (at least) four movements -- all of which target the bottom layers of the tree -the surface order can be derived is as follows:

- 10 -

(41)

Some orders of prefixes in Bulgarian instantiate the mirror-image of the fseq in (40) while some other orders instantiate (40) on the surface, eg: • repetitive pre follows excessive raz and cumulative na in both the surface and the basegenerated form • subsequence na > raz > pre follows the terminative do and completive iz only in the surface form in (39) • both do and iz preserve the initial relative order in the surface form. Conclusion about fseq in (3): • realized as (what is conceived of as) "morphology" and syntax • spelled out in Polish in the order it is base generated (without roll-up) • spelled out in Japanese in a mirror fashion (roll-up) while it remains spelled out in Polish (i.e. spelled out in Bulgarian follwing only a partial roll-up

Morphemes and complex constituents as heads of the same fseq fseq (like in (3), or any other) can be realized as prefixes (Slavic) or verbs (Japanese) in a system which allows for both terminal nodes but also complex constituents to project the label for the merger -- in particular a system like in Starke (2004): (42) morphemes as heads

(43) complex constituents as heads

- 11 -

References Aboh, Enoch. 2003. Morpho-Syntax of Head-Complement Sequences. Oxford: Oxford University Press. Baker, Mark. 1988. Incorporation: A Theory of Grammatical Function Changing. Illinois: University of Chicago Press. Caha, Pavel. 2009. The nanosyntax of case. PhD dissertation, University of Tromsø/CASTL. Cinque, Guglielmo. 1999. Adverbs and Functional Heads: A Cross-linguistic Perspective. New York: Oxford University Press. Cinque, Guglielmo. 2006. Restructuring and Functional Heads. Oxford: Oxford University Press. DiSciullo, Anna Maria, and Roumyana Slabakova. 2005. Quantification and aspect. In Perspectives on aspect, ed. by A. van Hout, H. de Swart, and H. Verkuyl, 61–80. Dordrecht: Springer. Endo, Yoshio. 2007. Locality and information structure: A cartographic approach to Japanese. Amsterdam/Philadelphia: John Benjamins Publishing Company. Istratkova, Vyara. 2004. On multiple prefixation in Bulgarian. Nordlyd 32,2: 301-321. Jabłońska, Patrycja. 2007. Radical decomposition and argument structure. PhD dissertation, University of Tromsø. Kageyama, Taro. 1993. Bunpoo to Gokeisei [Grammar and Word Formation]. Tokyo: Hituzi Syobo. Kageyama, Taro. 2009. Gengo no kozo seiyaku to jojutu kino [Structural constraints and predication functions in language] Gengo Kenkyu. 136: 1-34. Koopman, Hilda. 2005. Korean (and Japanese) morphology from a syntactic perspective. Linguistic Inquiry 36,4:601-633 Matsumoto, Yo. 1996. Complex Predicates in Japanese. Stanford: CSLI Publications. Nishiyama, Kunio. 2008. V-V compounds. Handbook of Japanese Linguistics, ed. by Shigeru Miyagawa and Mamoru Saito, pp.320-347. New York: Oxford University Press. Ramchand, Gillian. 2004. Time and the event: the semantics of Russian prefixes. Nordlyd 32:323–361. Romanova, Eugenia. 2004. Superlexical vs. lexical prefixes. Nordlyd 32,2:255–278. Starke, Michal. 2004. On the inexistence of specifiers and the nature of heads. In Adriana Belletti, ed. Structures and Beyond: The Cartography of Syntactic Structures, vol. 3, Oxford: Oxford University Press. Svenonius, Peter. 2004a (ed.). Nordlyd 32,2: Special Issue on Slavic Prefixes. Svenonius, Peter. 2004b. Slavic prefixes inside and outside VP. Nordlyd 32,2:205–253. Wiland, Bartosz. 2009. Aspects of order preservation in Polish and English. PhD dissertation, University of Poznań. Wiland, Bartosz. 2012. Prefix stacking, syncretism and the syntactic hierarchy. In Slavic Languages in Formal Grammar. Proceedings of FDSL 8.5, Brno 2010, ed. by Marketa Zikova and Mojmir Docekal, 307-324. Berlin/Frankfurt am Main: Peter Lang. Yumoto, Yoko. 2005. Hukugoo-Doosi, Hasei-Doosi-no Imi-to Toogo [The Semantics and Syntax of Compound Verbs and Derived Verbs]. Tokyo: Hituzi Shobo.

- 12 -